* To run the experiment, execute the following command:
	python run_experiment.py <[Jena | Pellet]> ./sample.json ./sample <result file>

* Data documents and the inference results are stored in ./sample directory.

* Each document in the sample is represented with an MD5 ID. The directory
 "sample/" contains all the analyzed data documents. Usually the processed
 results of the inference process are also stored in a file named
 "inputOnly.nq".  Furthermore, the triples in each classified derivation types
 (type1-type4) are also stored in the files that correspond to the derivation
 type. If there is no file "inputOnly.nq" in a directory of a data document, or
 files corresponding to "type1 - type4" are missing, that is because the
 inference process for the data document has failed.